Switchable Microtopographies
based on the Two-Way Shape Memory Effect
in Nickel-Titanium Alloys
Dissertation
zur Erlangung des Grades
des Doktors der Ingenieurwissenschaften der Naturwissenschaftlich-Technischen Fakultät III Chemie, Pharmazie, Bio- und Werkstoffwissenschaften
der Universität des Saarlandes
von
MAREIKE FRENSEMEIER
Angefertigt am INM – Leibniz Institut für Neue Materialien Metallische Mikrostrukturen / Schaltbare Oberflächen
Saarbrücken, Januar 2016
Tag des Kolloquiums: 24. Juni 2016
Dekan: Prof. Dr.-Ing. Dirk Bähre Vorsitzender: Prof. Dr.-Ing. Frank Mücklich Berichterstatter: Prof. Dr. Eduard Arzt
Prof. Dr. Stefan Seelecke Prof. Dr. Anand Jagota Akad. Mitarbeiter: Dr.-Ing. Frank Aubertin
Meinen Eltern
“The greatest compliment that was ever paid me was
when one asked me what I thought,
and attended to my answer.”
Nickel-titanium (NiTi) shape memory alloys are functional materials that are capable of undergoing a reversible temperature-induced shape change. Specifically in martensitic NiTi alloys, a reversible two-way shape memory effect can be induced using indentation techniques enabling a temperature-induced change in topography. Combining switchable topographies with nano- or microstructures could expand the properties of functional surfaces, and in addition make the surfaces responsive to their environment. For example, it would be possible to change the adhesive properties of surfaces with switchable dry adhesive microstructures or to control celladhesion on implant materials with specific nano- and micro-switchable structures.
In this study, the indentation induced two-way shape memory effect was investigated in different NiTi alloys. In particular, the effects of alloy microstructure, deformation parameters (training) and thermal treatments on switchability were explored. In an austenitic NiTi alloy a specific thermal treatment led to the formation of coherent precipitates, which were shown to be crucial for the two-way shape memory behavior; exceeding the phase transformation temperature considerably decreased the switchability of the topography. At higher temperatures the stabilized martensite, which is required for an oriented phase transformation and consequently for the two-way shape memory behavior, transforms to austenite.
An embossing and an electrochemical forming process were developed to prepare switchable topographies on larger areas. Both methods led to surface arrays on NiTi with two-way shape memory topographies.
Finally, two approaches were presented, which use the switchable topographies to enable switching of a formerly passive surface function. In combination with bioinspired dry adhesive structures, the switchable NiTi topography led to a reversible, temperature-induced change of the adhesive properties of the surface. Secondly, the two-way shape memory effect was transferred to an alloy system with a phase transformation temperature near body temperature and a small width of hysteresis. By this, a switchable topography was induced, which is controllable within a physiological temperature range. The only issue impeding the use of this switchable surface for experiments on cell-surface interactions is an increased leakage of harmful copper ions. Therefore, surface passivation through oxidation is presented as a method to reduce the ion leakage.
Kurzzusammenfassung
Nickel Titan (NiTi) Formgedächtnislegierungen ermöglichen eine Formänderung als Funktion der Temperatur. Bei martensitischen NiTi-Legierungen kann durch Indentieren der Oberfläche ein reversibler Zwei-Weg Formgedächtniseffekt induziert werden, der die Ausbildung thermisch schaltbarer Topographien bewirkt. Die Kombination dieser schaltbaren Topographie mit funktionalen Nano- oder Mikrostrukturen erlaubt die Entwicklung von Oberflächen mit schaltbaren Eigenschaften. Beispielsweise ließe sich durch eine schaltbare Topographie das Zellwachstum auf Implantatmaterialien steuern, oder in Kombination mit trocken-adhäsiven Mikrostrukturen, die Adhäsionseigenschaften einer Oberfläche regulieren.
In der vorliegenden Arbeit wurde der indentations-induzierte Zwei-Weg Formgedächtniseffekt in verschiedenen Legierungen untersucht, insbesondere der Zusammenhang zwischen Schaltbarkeit und Verformung, Mikrostruktur und Temperaturänderung. In einer austenitischen NiTi-Legierung führte eine gezielte Temperaturbehandlung zur Bildung kohärenter Ausscheidungen, welche sich als entscheidend für eine schaltbare Topographie gezeigt haben. Wurde die Phasenumwandlungstemperatur weit überschritten, wandelte sich der für den Zwei-Weg Formgedächtniseffekt notwendige stabilisierte Martensit in Austenit um, was eine Rückumwandlung der Topographie verhindert.
Um schaltbare Topographien auf größeren Flächen zu erzeugen, wurden ein Kaltpräge- und ein elektrochemisches Abtragungsverfahren entwickelt. Beide Methoden ermöglichten die Herstellung von schaltbaren NiTi- Oberflächen.
Dass dynamische Topographien in NiTi für die Schaltbarkeit einer passiven Oberflächenfunktion genutzt werden können, wurde in zwei Ansätzen aufgezeigt. In Kombination mit bioinspirierten Haftstrukturen wurde eine reversible temperatur-induzierte Änderung der adhäsiven Eigenschaften einer Oberfläche ermöglicht. Außerdem wurde der Zwei-Weg Formgedächtniseffekt auf ein quaternäres Legierungssystem mit einer Phasenumwandlungstemperatur nahe der Körpertemperatur und schmaler Hysteresebreite übertragen. Hieraus resultierte eine Oberfläche, deren Topographie sich innerhalb eines physiologisch tolerierbaren Temperaturfensters schalten lässt. Einzig eine erhöhte Abgabe von zellschädigenden Kupferionen stand der Untersuchung einer Zellwechselwirkung mit der schaltbaren Oberfläche entgegen. Mit dem Ziel die Ionenabgabe zu reduzieren, wurde deshalb die Oberfläche durch Oxidation passiviert.
Danksagung
Die vorliegende Dissertation ist am INM – Leibniz Institut für Neue Materialien in Saarbrücken unter der Leitung von Prof. Dr. Eduard Arzt in den Forschungsgruppen metallische Mikrostrukturen und schaltbare Oberflächen entstanden.
Während der Bearbeitung habe ich von vielen Seiten, wissenschaftlich wie privat, Unterstützung erfahren dürfen. An dieser Stelle möchte ich mich bei allen für ihren Beitrag zu dieser Arbeit bedanken.
Allen voran gilt mein Dank meinem Doktorvater Prof. Dr. Eduard Arzt für die Ermöglichung dieser Dissertation, die konstruktiven Anregungen und die stetige Unterstützung.
Ebenfalls möchte ich mich herzlich bei Prof. Dr. Stefan Seelecke für die motivierenden Gespräche und die Übernahme des Zweitgutachtens bedanken.
Für ihren Einsatz während und nach ihrer Zeit als Juniorforschungsgruppenleiter am INM verdienen Dr. Andreas Schneider und Dr. Elmar Kroner meine Anerkennung und ein großes Lob. Mit ihren fachlichen Ratschlägen und neuen Ideen haben sie einen besonderen Beitrag zum Fortgang dieser Arbeit geleistet. Neben einem freundschaftlichen Arbeitsklima stieß ich bei ihnen stets auf ein offenes Ohr, viel Verständnis und, nicht zu vergessen, eine konstante Versorgung mit Nervennahrung. Weiterhin möchte ich auch Dr. Carl Frick von der University Wyoming danken, er hat mich nicht nur mit den komplexen Eigenschaften von Formgedächtnislegierungen vertraut gemacht sondern auch damit, dass Wyoming deutlich mehr als Cowboys zu bieten hat.
Ein ‚Dangschee‘ gilt auch Birgit Heiland - ohne die vielen Tipps und Tricks zur Probenpräparation ‚däd i wohl heid no poliera ond schloifa ond häd nichts gmachd ghedd‘. Ebenso möchte ich mich bei Jörg Schmauch, Rudolf Karos, Robert Drumm und
Dr. Yuliya Silina für ihre Unterstützung bei den TEM Aufnahmen, XRD-, DSC und ICP-MS Messungen bedanken sowie bei Dr. Vera Bandmann und Isabella Reichert für die Durchführung der Zellversuche. Für die Konstruktion und kontinuierliche Rekonstruktion des Mikro-Heizers gilt mein Dank dem Team der Werkstatt, das sich den Verbesserungswünschen und Reparaturen geduldig angenommen hat.
Dank des großen Zusammenhalts und der Unternehmungsfreude des gesamten Kollegenkreises kam auch der Ausgleich abseits der Wissenschaft nicht zu kurz. Neben
Jäckel, Jessica Kaiser, Bentejui Medina Clavijo, Maurizio Micciché, Nicolas Peter, Julia Purtov, Dominik Schirra, Mariana Viegas Greco de Oliveira und Paula Yagüe Isla für das schöne Gefühl bedanken, mit guten Freunden zusammen zu arbeiten.
Abschließend gilt mein persönlicher Dank meinen Freunden und meiner Familie in der Heimat sowie meinem Freund Jona. DANKE, für die großartige Unterstützung trotz der zeitweise großen Entfernung, die Ausdauer und das entgegengebrachte Verständnis während der Anfertigung dieser Arbeit.
Abbreviations and symbols
A Contact area
a Radius of the contact area
Af Austenite finish temperature
As Austenite start temperature
as Aspect ratio of the PECM structures
d Diameter of pillar tip
dc Indentation depth after cooling below Mf
dh Indendation depth after heating above Af
di Indentation depth after indentation
DSC Differential scanning calorimetry
E Young’s modulus
E* Reduced elastic modulus
EBSD Electron backscatter diffraction ECM Electrochemical machining EDM Electrodischarge machining
F Applied load
FC Pull-off force, minimum negative load
FCB Pull-off force in the bumpy state
FCF Pull-off force in the flat state
FEA Finite element analysis FIB Focused ion beam
hs Height of the structures
∆ Enthalpy difference
I Current pulse
ICDD International center for diffraction database ICP-MS Inductively coupled plasma mass spectrometry JKR Johnson-Kendall-Roberts modell
lc Diameter of the cavity in the cathode
lN Liquid nitrogen
ls Length of cross-section of the structures at the base
OWSME One-way shape memory effect Ratio of load to pull-off force PBS Phosphate-buffered saline PDMS Polydimethylsiloxane
PECM Pulse electrochemical machining
R Reduced radius
Ra Average roughness
RCF Effective radius in the flat state
RCB Effective radius in the bumpy state
Rf R-phase finish temperature
RROW Recovery ratio one-way shape memory
Rs R-phase start temperature
RRTW Recovery ratio two-way shape memory
Rz Peak-to-valley roughness
S Switching efficiency ∆ Entropy difference
SEM Scanning electron microscopy SIM Stress induced martensite SMA Shape memory alloy SME Shape memory effect
T Stress-induced transformation temperature
T0 Temperature of thermodynamic equilibrium
TEM Transmission electron microscopy TWSME Two-way shape memory effect WLI White-light interferometer XRD X-Ray Diffraction
w Work of adhesion
z0 Equilibrium separation
ε Strain
σy Yield stress
υ Poisson’s ratio
Chapter 1
Introduction
Functional surfaces with super-hydrophobic, (anti)-reflective, adhesive or bioactive properties play a central role for new developments in research areas such as tribology, microfluidics or biomedicine.1-5 In many cases these surface properties were induced
using functional micro- or nanostructures or even a hierarchical morphology, as found in the lotus leaf, the natural prototype for super-hydrophobic surfaces.6 In addition,
functional coatings and changes in the surface texture are frequently used to improve or change surface properties such as wettability or friction.7-8 For these properties the
surface topography on micro- and nanometer scale plays an important role and induces macroscopic effects.9
So far, mostly passive structures and materials have been used to obtain certain functionalities. Recently, investigations in the field of ‘smart’, ‘stimuli-responsive’, ‘intelligent’, ‘active’ or ‘dynamic’ materials and hybrid systems have gained increasing attention. The ability to change and control specific surface properties could enable applications from data-storage, over microelectromechanical systems to dynamic biomedical surfaces, guiding cell-adhesion, and surfaces with adaptive reflectivity. Besides complex molecular systems, which show a responsive chemical modification, shape memory materials, which are able to manifest a shape change, have been used to trigger surface functionalities. Many of these systems are based on polymers changing their physio-chemical properties activated by pH, humidity, light or temperature.10-12 Only a few studies are available using inorganic and metal-based
responsive surfaces. Most of these switchable systems obtain their active properties by a combination with a modified polymeric surface layer or physical adsorption of polymer molecules.12-13 These hybrid systems combine the mechanical advantages of a passive
metal with the switchable properties of an active polymer. However, switchable metallic topographies could surpass switchable polymeric topographies in particular for
load bearing or high temperature applications. Magnetism, electric currents and inductive heating could be used as external stimulus and induce a shape change in shape memory alloys with actuation forces exceeding those of polymers by far.14-16
Regarding biomedical applications, metal based implants were frequently used and medically proven.17
Prominent responsive metallic materials are nickel-titanium (NiTi) shape memory alloys. Near equiatomic NiTi alloys show a reversible shape change triggered by a change in temperature and extraordinarily high elastic deformation, called pseudoelasticity. Their exceptional mechanical properties make them a frequently used material in actuator applications and biomedicine.18-19 For example, pseudoelastic NiTi
materials are used in orthodontic wires or in flexible stents for endovascular therapy. 20-22 Only few applications, such as actuators, control devices, or fasteners, make use of
the two-way shape memory behavior in NiTi alloys.23-25
Depending on the alloy composition and specific thermomechanical training, either the one-way shape memory effect (OWSME) or the two-way shape memory effect (TWSME) can be exploited. Previous research on the two-way shape memory effect has established temperature-induced switchable topographies in NiTi shape memory alloys. Indentation followed by planarization yielded a shape memory effect, where the topography of the NiTi surface reversibly changed from flat to textured. The switch is controlled by heating above, or cooling below the characteristic phase transformation temperatures of the alloy.16, 26-28
So far, the detailed microstructural mechanisms behind the indentation induced TWSME remain unclear, especially the influence of heat treatments on the TWSME is not well understood. A better knowledge on the adaptation of the phase transformation temperatures and on the enlargement of protrusion height and reversibility is required for technical application. For example, could this enable a switch of topographies at physiological conditions and improve the manufacturing technology for such switchable metallic topographies.
Apart from a study by Fei et al. the focus of previous studies has been on the scientific investigation of the effect. The same indentation preparation technique presented in the first study by Zhang et al. was used, which is rather complex and time consuming.27, 29 No method has been presented for a training of the TWSME on larger
areas that enables a parallel preparation. This step is crucial for upscaling of switchable metallic arrays and subsequent implementation into process engineering.
Apart from a wear resistant coating, applications of metallic switchable topographies have been rarely investigated. 30 The usage of such a switchable metallic
topography is very promising regarding the control of surface properties. The change in topography from flat to textured might have a strong impact on the contact area. In particular, properties which are strongly related to the surface topography, such as friction, wettability or adhesion, could be controlled.
The development of a hybrid system, combining switchable metallic surface topographies with functional coatings and microstructures could further extend the field of applications. For example, could a combination with dryadhesive microstructures induce switchable adhesive properties for pick and place applications. In biomedicine, the control of the surface topography of a hybrid NiTi implant system may promote the healing process, e.g. by controlled drug release or by influencing cell proliferation due to controlled wettability and surface roughness.31-32
With respect to these possible applications, the aim of this thesis is to understand the indentation induced shape memory effect and its influence on switchable surface structures in more detail. Furthermore, with a view to a possible upscaling, preparation techniques that enable large area patterning were developed. Finally, two applications of switchable metallic topographies will be presented; an approach using a shape memory topography in a NiTi-polymer hybrid system to gain temperature-induced switchable adhesion, and a switchable topography activated near body temperature using a new shape memory alloy. This thesis is outlined as follows:
After the introduction in the first chapter, in the second chapter the nature of NiTi shape memory alloys, the martensitic phase transformation and the indentation induced two-way shape memory effect are described. Furthermore, a brief overview on the current state of the art in micro manufacturing and structuring of NiTi alloys is given. The chapter ends with a detailed overview of the current research in the field of functional micro- and nanostructures, coatings and topographies. In particular, developments on switchable dry adhesive and biomedical surfaces are emphasized.
The third and the fourth chapter present the research on the influence of precipitates on the two-way shape memory effect. This work was started by Dr. Enwei Qin and completed during this thesis with additional differential scanning calorimetry (DSC), indentation and temperature cycling measurements to analyze the effect of aging on the indentation induced two-way-shape memory effect in austenitic NiTi. In
to train switchable metallic surfaces based on the TWSME. Three different geometries are induced by these techniques. The resulting switchable surface arrays are quantified with focus on their morphology and reversibility using white light interferometry and thermal cycling. Chapter 6 discusses the development of a hybrid system by combining a switchable metallic topography in NiTi with a bioinspired, micro-patterned adhesive polymer layer. The adhesion properties and reversibility of the hybrid system are measured as a function of temperature. The results are discussed in terms of a change in contact area using classic contact mechanics and a finite element analysis.
Chapter 7 introduces switchable metallic topographies activated near body temperature using a quaternary shape memory alloy. The reversible switch between flat and structured is used to investigate the control of cell adhesion and cell alignment without chemical surface modifications.
The thesis concludes in Chapter 8 with a general discussion of the most relevant results presented in this work, and an outlook on possible investigations based on the presented findings is given. Further, ideas for subsequent developments using metallic shape memory topographies are introduced.
1.1 Publications and author contributions
Parts of the present thesis have been published with contributions of co-authors in refereed scientific journals. Two manuscripts were mainly contributed by the author of this thesis and published with first authorship; the third publication was published in co-authorship.
» Chapter 3 was published by Qin, E.; Peter, N. J.; Frensemeier, M.; Frick, C. P.; Arzt, E.; Schneider, A. S. in the article Vickers Indentation Induced One-Way and Two-Way Shape Memory Effect in Austenitic NiTi, Adv. Eng. Mater. 2014, 16 (1), 72-79.
The author carried out the indent-depth-measurements and the evaluation and interpretation of the results. The writing of the paper was conducted in close collaboration with C. Frick and N. Peter. E. Qin was responsible for the sample heat treatment and TEM measurements. The project was supervised by E. Arzt and A. S. Schneider.
» Chapter 4 was published by Frensemeier, M.; Arzt, E.; Qin, E.; Frick, C. P.;
in Aged Ti-50.9 at.% Ni, in MRS Commun. 2015, 5 (01), 77-82.
The first author was responsible for the conduction and evaluation of the DSC, topography and temperature cycling measurements. The TEM imaging and heat treatment was carried out by E. Qin. The author wrote the manuscript and received scientific supported by E. Arzt and C. P. Frick. Both contributed with fruitful discussions. The idea and concept of the publication was developed by A. S. Schneider.
» Chapter 5 was published by Frensemeier, M., Schirra, D., Weinmann, M., Weber, O., Kroner, E. in the article Shape-Memory Topographies on Nickel-Titanium Alloys Trained by Embossing and Pulse Electrochemical Machining in Adv. Eng. Mater. 2016, doi: 10.1002/adem.201600012
The author carried out the concept of the work, the sample preparation and topography change measurements. The writing of the paper was conducted in close collaboration with E. Kroner. D. Schirra performed the reversibility tests under supervision of M. Frensemeier. M. Weinmann and O. Weber provided the PECM system and were responsible for the PECM structuring process.
» Chapter 6 was published in the article Temperature-Induced Switchable Adhesion using Nickel–Titanium–Polydimethylsiloxane Hybrid Surfaces, by Frensemeier, M.; Kaiser, J. S.; Frick, C. P.; Schneider, A. S.; Arzt, E.; Fertig, R. S.; Kroner, E. in
Adv. Funct. Mater. 2015, 25 (20), 3013-3021.
The first author was responsible for all laboratory work and measurements, particularly the establishment of the hybrid systems. The adhesion measurements were supported by J. Kaiser and the advice from E. Kroner. The FEM analysis was performed by R.S. Fertig. The author has coordinated the summary of the results and the writing of the manuscript. The final manuscript was discussed and corrected with the suggestions from the co-authors C.P. Frick, E. Arzt and A.S. Schneider. The project was supervised by E. Arzt and E. Kroner.
1.2 References
1. Feng, L.; Li, S.; Li, Y.; Li, H.; Zhang, L.; Zhai, J.; Song, Y.; Liu, B.; Jiang, L.; Zhu, D., Super-hydrophobic surfaces: from natural to artificial. Advanced Materials
2002, 14 (24), 1857-1860.
2. Li, Y.; Zhang, J.; Yang, B., Antireflective surfaces based on biomimetic nanopillared arrays. Nano Today 2010, 5 (2), 117-127.
3. Kikuta, H.; Toyota, H.; Yu, W., Optical elements with subwavelength structured surfaces. Optical Review 2003, 10 (2), 63-73.
4. Zhao, G.; Raines, A.; Wieland, M.; Schwartz, Z.; Boyan, B., Requirement for both micron-and submicron scale structure for synergistic responses of osteoblasts to substrate surface energy and topography. Biomaterials 2007, 28 (18), 2821-2829. 5. Wennerberg, A.; Albrektsson, T.; Andersson, B.; Krol, J., A histomorghometric study
of screw-shaped and removal torque titanium implants with three different surface topographies. Clinical oral implants research 1995, 6 (1), 24-30.
6. Koch, K.; Bhushan, B.; Jung, Y. C.; Barthlott, W., Fabrication of artificial Lotus leaves and significance of hierarchical structure for superhydrophobicity and low adhesion. Soft Matter 2009, 5 (7), 1386-1393.
7. Ponsonnet, L.; Reybier, K.; Jaffrezic, N.; Comte, V.; Lagneau, C.; Lissac, M.; Martelet, C., Relationship between surface properties (roughness, wettability) of titanium and titanium alloys and cell behaviour. Materials Science and Engineering: C 2003, 23 (4), 551-560.
8. Pettersson, U.; Jacobson, S., Influence of surface texture on boundary lubricated sliding contacts. Tribology International 2003, 36 (11), 857-864.
9. Assender, H.; Bliznyuk, V.; Porfyrakis, K., How Surface Topography Relates to Materials' Properties. Science 2002, 297 (5583), 973-976.
10. Ichimura, K.; Oh, S.-K.; Nakagawa, M., Light-Driven Motion of Liquids on a Photoresponsive Surface. Science 2000, 288 (5471), 1624-1626.
11. Berna, J.; Leigh, D. A.; Lubomska, M.; Mendoza, S. M.; Pérez, E. M.; Rudolf, P.; Teobaldi, G.; Zerbetto, F., Macroscopic transport by synthetic molecular machines.
Nature materials 2005, 4 (9), 704-710.
12. Skorb, E. V.; Andreeva, D. V., Surface Nanoarchitecture for Bio-Applications: Self-Regulating Intelligent Interfaces. Advanced Functional Materials 2013, 23 (36), 4483-4506.
13. Chen, M.; Yang, X.; Hu, R.; Cui, Z.; Man, H., Bioactive NiTi shape memory alloy used as bone bonding implants. Materials Science and Engineering: C 2004, 24 (4), 497-502.
14. Müller, C. W.; Pfeifer, R.; El-Kashef, T.; Hurschler, C.; Herzog, D.; Oszwald, M.; Haasper, C.; Krettek, C.; Gösling, T., Electromagnetic induction heating of an orthopaedic nickel–titanium shape memory device. Journal of Orthopaedic Research 2010, 28 (12), 1671-1676.
15. Ohandley, R. C.; Murray, S. J.; Marioni, M.; Nembach, H.; Allen, S. M., Phenomenology of giant magnetic-field-induced strain in ferromagnetic shape-memory materials (invited). Journal of Applied Physics 2000, 87 (9), 4712-4717. 16. Fei, X. L.; Zhang, Y. J.; Grummon, D. S.; Cheng, Y. T., Indentation-induced two-way
shape memory surfaces. Journal of Materials Research 2009, 24 (3), 823-830. 17. Park, J. B.; Lakes, R. S., Metallic implant materials. Biomaterials 2007, 99-137. 18. Duerig, T.; Pelton, A.; Stöckel, D., An overview of nitinol medical applications.
19. Saadat, S.; Salichs, J.; Noori, M.; Hou, Z.; Davoodi, H.; Bar-On, I.; Suzuki, Y.; Masuda, A., An overview of vibration and seismic applications of NiTi shape memory alloy. Smart Materials and Structures 2002, 11 (2), 218.
20. Blum, U.; Voshage, G.; Lammer, J.; Beyersdorf, F.; Töllner, D.; Kretschmer, G.; Spillner, G.; Polterauer, P.; Nagel, G.; Hölzenbein, T., Endoluminal stent–grafts for infrarenal abdominal aortic aneurysms. New England Journal of Medicine 1997,
336 (1), 13-20.
21. El Feninat, F.; Laroche, G.; Fiset, M.; Mantovani, D., Shape memory materials for biomedical applications. Advanced Engineering Materials 2002, 4 (3), 91.
22. Miura, F.; Mogi, M.; Ohura, Y.; Hamanaka, H., The super-elastic property of the Japanese NiTi alloy wire for use in orthodontics. American Journal of Orthodontics and Dentofacial Orthopedics 1986, 90 (1), 1-10.
23. Pfeifer, R.; Müller, C. W.; Hurschler, C.; Kaierle, S.; Wesling, V.; Haferkamp, H., Adaptable orthopedic shape memory implants. Procedia CIRP 2013, 5, 253-258. 24. Kim, H.-C.; Yoo, Y.-I.; Lee, J.-J., Development of a NiTi actuator using a two-way
shape memory effect induced by compressive loading cycles. Sensors and Actuators A: Physical 2008, 148 (2), 437-442.
25. Perkins, J., Shape memory effects in alloys. Springer Science & Business Media: 2012.
26. Ni, W.; Cheng, Y.-T.; Grummon, D. S., Microscopic shape memory and superelastic effects under complex loading conditions. Surface and Coatings Technology 2004,
177, 512-517.
27. Zhang, Y.; Cheng, Y. T.; Grummon, D. S., Shape memory surfaces. Applied Physics Letters 2006, 89, 041912.
28. Zhang, Y.; Cheng, Y.-T.; Grummon, D. S., Two-way indent depth recovery in a NiTi shape memory alloy. Applied Physics Letters 2006, 88 (13), 131904-131904-3. 29. Fei, X.; Grummon, D. S.; Ye, C.; Cheng, G. J.; Cheng, Y.-T., Surface form memory
in NiTi shape memory alloys by laser shock indentation. Journal of Materials Science 2012, 47 (5), 2088-2094.
30. Ni, W.; Cheng, Y.-T.; Grummon, D. S., Wear resistant self-healing tribological surfaces by using hard coatings on NiTi shape memory alloys. Surface and Coatings Technology 2006, 201 (3), 1053-1057.
31. Hermawan, H.; Dubé, D.; Mantovani, D., Developments in metallic biodegradable stents. Acta Biomaterialia 2010, 6 (5), 1693-1697.
32. Ranella, A.; Barberoglou, M.; Bakogianni, S.; Fotakis, C.; Stratakis, E., Tuning cell adhesion by controlling the roughness and wettability of 3D micro/nano silicon structures. Acta Biomaterialia 2010, 6 (7), 2711-2720.
Chapter 2
Fundamentals and literature review
2.1 Shape memory effects in nickel-titanium alloys
Intermetallic NiTi shape memory alloys respond to external mechano-caloric stimuli. They are frequently used in micromechanical and biomedical devices as actuators, sensors and implant materials.1-3 All of these applications originate from the
shape memory effect or pseudoelasticity; two complex behaviors associated with an austenite-martensite phase transformation.4 In order to increase the magnitude of
shape change and the reversibility, the underlying martensitic phase transformation has been the subject of continued research since its discovery in 1963.5 NiTi alloys show
extraordinarily strong pseudoelastic properties and a reversible shape change even for relatively large strains up to 8%.6 In addition to their shape memory behavior, good
corrosion resistance and biocompatibility compared to other shape memory alloys makes them an excellent candidate for advanced biomedical and implant applications. 7-8
2.1.1 The martensitic phase transformation
NiTi alloys with shape memory effect typically have a near equiatomic composition. At moderate temperatures (<650 °C), the region of homogeneity is quite narrow and lies between 50.0 and 50.5 at.% Ni.9 In this temperature range, slightly Ni-rich alloys
(>50.5 at.% Ni) decompose upon ageing into metastable Ti3Ni4, Ti2Ni3 and subsequently
TiNi3. Especially, the Ti3Ni4 phase influences the shape memory effect due to a lattice
distortion by the formation of coherent precipitates.10-11 Their anisotropic shrinking of
2.7 % in the [111]-direction and only 0.3 % in the perpendicular orientation leads to a lenticular shape surrounded by strain fields, as shown in Figure 2.1.11 The phase
alloying elements, and by thermo-mechanical treatments of the alloy. In binary NiTi alloys of 50.0 to 51.5 at.% Ni, Ms decreases by 150°C per at.% Ni until no martensite
phase transformation occurs anymore.12
The shape change recovery of NiTi shape memory alloys is based on a martensitic phase transformation, a shear-dominant diffusionless solid-state phase transformation. As shown in Figure 2.2, the martensitic phase transformation involves an athermal first-order displacive transformation from the cubic B2 high-temperature parent phase (austenite, A) to the monoclinic B19’ low-temperature phase (martensite, B).4 Ageing
and thermo-mechanical treatments, enable a two-step transformation from the highly symmetric B2 phase over the trigonal R-Phase to the lower symmetric B19’ phase.13
The martensitic B19’ unit cell is elongated over 10% compared to the austenitic B2 unit cell in the [223] B2-orientation and an order larger than the R-phase with an elongation of 0.94% along the [111] B2-orientation.14-16 The shape change in NiTi alloys
is driven by this change of transformational volume and the high mobility of twin boundaries under the influence of shear stress. In polycrystalline NiTi shape memory alloys, the martensitic phase transformation is driven by nucleation and growth. Nucleation preferably evolves at phase and grain boundaries, at temperatures below the martensitic start temperature Ms. Below the martensitic finish temperature Mf the
phase transformation is completed and the material is fully martensitic. Analogously, the transformation from the low temperature martensitic phase to the high temperature austenitic phase starts at the austenite start temperature As and
continuous until the austenite finish temperature Af is reached and the material is in
a fully austenitic state.
Figure 2.1: Schematic of the lattice distortion and intrinsic stress fields induced by shrinking of coherent Ni4Ti3 precipitates. Modified after Tadaki et al.11
Ti3Ni4 [111]B2
T > 680°C
The phase transformation temperatures can be altered by changing the composition, additional alloying elements, and by thermo-mechanical treatments of the alloy. In binary NiTi alloys of 50.0 to 51.5 at.% Ni, Ms decreases by 150°C per at.% Ni
until no martensite phase transformation occurs anymore.12 Moreover, the formation of
precipitates in Ni-rich alloys and the formation of a two-step transformation due to the R-Phase have to be considered.17
Phase transformation temperatures are usually determined by DSC, and range for NiTi from -150 to 100 °C. This temperature range generally defines the field of application of the shape memory alloy.
Furthermore, phases can be stabilized by mechanical stress. Due to the displacive nature of the phase transformation, martensite can also be induced by stress (SIM). The correlation between induced stress and temperature is described by the Clausius-Clapeyron equation:18
dT
dσ= −∆ = −∆
(2.1)
Here, T is the stress-induced transformation temperature, is the transformation induced strain, ∆ is the entropy, ∆ the enthalpy difference of the phase transformation, and T0 the temperature of thermodynamic equilibrium. Accordingly,
martensite can be induced by mechanical stress at a constant temperature (T > Af). It
spontaneously recovers after unloading since the reverse transformation takes place.
Figure 2.2: Sketch of the crystal structures in NiTi shape memory alloys.4, 19 (A) The parent
austenitic B2 structure, with four primitive cells leading to a tetragonal unit cell (dashed line). (B), The phase transformation from austenite (B2) to martensite (B19’) is induced by shifting the (110) plane in the direction and a cooperative shearing of the (001) planes in [110]-direction.
This stress-induced reversible deformation, called pseudoelasticity, is limited by the temperature dependent critical stress for plastic yielding in the austenitic phase (Figure 2.3).
Once the applied stress level overcomes the critical stress for slip, before reaching the level for a stress-induced martensitic transformation, permanent deformation occurs.20Ms is shifted to higher temperatures with increasing stress, until the critical
yield stress, σy, is reached. With further increasing stress level, the pseudoelastic
deformation is displaced by plastic deformation.
2.1.2 The shape memory behavior
Depending on the thermomechanical treatment and loading path, the diffusion-less stress-induced martensitic transformation leads to either pseudoelastic, one-way shape memory or a two-way, i.e. repeatable, shape memory behavior (Figure 2.4). At temperatures below Mf, two different variations of martensite accommodation are
shown.
Figure 2.3: Schematic relationship between temperature and stress-induced martensitic phase transformation. After Lagoudas et al.20
If the martensitic transformation takes place in absence of any external mechanical stimuli, e.g. an applied stress, the multiple variants of the low-symmetric martensite show an intrinsic self-accommodation due to twinning. In this case, no macroscopic shape change can be observed. The change in symmetry enables the formation of twin-related pairs of the martensite variants, which nucleate to a single, coherent habit-plane variant.
If the transformation takes place in presence of an external mechanical load, the multiple martensitic variants re-accommodate with a preferred orientation (Figure 2.4). This detwinning process leads to a large, reversible (non-plastic) strain, accompanied by a macroscopic shape change. After heating above Af, the material
returns to the highly symmetric, single variant austenite phase and the deformation fully recovers.
Figure 2.4: Schematic stress-strain diagram and martensitic phase transformation of (A) the one-way shape memory effect and (B) pseudoelasticity.
The memorized shape is always the austenitic shape, which can be recovered by heating the deformed, martensitic sample. Once the original austenitic shape is recovered, no re-deformation occurs during a subsequent cooling. Thus, this shape memory behavior is restricted to one switching process, the so-called one-way shape memory effect (OWSME). If the macroscopic shape recovery is inhibited during transformation from martensite to austenite, high restoring forces occur. Combined with an external bias force, the one-way shape memory effect can be used as a switchable actuator.21
For applications with operating temperatures above Af, another effect is elicited due
to mechanical loading, referred to as pseudoelasticity. If the material is deformed at temperatures near or above Af, a spontaneous stress induced formation of SIM takes
place. Similarly, to the transformation mechanism of the OWSME, favored martensite variants start growing and detwin depending on the magnitude of the applied stress. Thus, the microstructure and the macroscopic deformation is assumed to be the same as for lower temperatures. When the external load is removed, a sudden re-transformation to the austenitic shape occurs. The pseudoelastic deformation is directly related to the phase equilibrium temperature and the martensitic transformation temperature. It can be controlled in a mechanically loaded state and decreases immediately in the absence of load. The martensite recovers isothermally to highly symmetric austenite.
In order to elicit the two-way shape memory effect, an appropriate thermo-mechanical ‘training procedure’ has to be applied to the shape memory alloy.22-23 The
combination of deformation and heat treatment enables the material to memorize its shape at both high and low temperature. Thus, the two-way shape memory effect (TWSME) allows switching between the two shapes controlled by cycling temperature.
While the TWSME shows relatively low recovery strains compared to the OWSME, it shows high reversibility and high fatigue resistance.24-25 Most training procedures
create internal stress fields, which then guide the growth of oriented martensite variants. Usually, a thermomechanical training process is applied to the material, which consists of a defined heating and cooling procedure under controlled stress, strain and temperature.3, 26 The applied stress or strain often exceeds the recoverable strain
and leads to plastic slip and the generation of dislocations and stress fields. It is believed, that these stress fields act as nucleation sites for martensite variants with a
favored orientation, a process that is crucial for the shape deformation in absence of an external load (Figure 2.5).23 After the ‘training’ (i.-iii.), which induces a permanent
dislocation structure in the material, a reversible deformation occurs spontaneously during thermal cycling across the transformation temperatures without any external bias (iv. –vi.).27-28
The internal stress, generated by the dislocations, is associated with the stabilization of martensitic variants in favored orientations. During the thermally induced transformation, these pre-oriented variants guide the direction of the overall martensitic transformation leading to a macroscopic deformation of the sample.22
The reversible shape change due to the TWSME has been demonstrated in various studies.29-31 However, theoretical attempts explaining the underlying fundamental
principles of the TWSME were mostly based on the analysis of experimental phenomena. A detailed analysis of the interaction between deformation, associated dislocation structures, precipitates, stress fields, and the generation of oriented martensite variants is currently missing. Only a few studies investigated the specific interaction between deformation, precipitates and the activation of martensite growth with certain variant orientation.25, 32-33
Figure 2.5: Characteristic stress-strain behavior of the two- way shape memory effect (TWSME). The oriented microstructural change during deformation and the martensitic phase transformation lead to a macroscopic shape change induced by changing temperature. No external stress is required for the shape recovery.
For example, the interaction of the dislocation structure with favored martensitic variants was investigated by transmission electron microscopy (TEM) of fatigued NiTi samples by Gall and Maier.25 It was suggested that defect nucleation is energetically
favored at the austenite-martensite interface. Norfleet et al. showed that the amount of stress-induced martensite correlates with a high dislocation density and that the orientation of the martensite twins is parallel to the dislocation accumulation.34
Nishida and Honma, and Gall et al. showed that, in addition to the dislocation structure, the presence of precipitates may influence the behavior of the TWSME.10, 29
The phase transformation behavior (see 2.1.1) changes with internal stress fields. Therefore, size, amount, volume and orientation of precipitates were directly correlated to the growth of oriented martensite variants.33, 35 These studies have highlighted that
the guidance of oriented martensite growth by precipitates and dislocation structures is the key-element for TWSME evolution, a fundamental understanding of the TWSME is still missing.
2.1.3 The indentation induced two-way shape memory effect
While the fundamental mechanisms causing the reversible two-way shape transformation remain unclear, recent research has identified a new mechanism for inducing the TWSME. Besides relatively complex treatments such as thermo-mechanical cycling or stress-assisted aging, severe plastic deformation, e.g. through indentation, can induce a reversible shape memory effect.36
Studies by Ni et al. have shown that a TWSME can already be induced by a single indentation, without repeated thermal cycling or subsequent heat treatment. The indentation depth in a NiTi-surface was shown to decrease or increase upon heating and cooling above and below the transformation temperature, respectively. Thus, after indentation the material elicit a reversible shape memory effect, i.e. a TWSME (Figure 2.6).37
Most of the decrease in indentation depth is related to the OWSME and occurs after heating the sample above Af. After cooling the sample below Mf, the indent depth
increases again, leading to a partially reverse deformation related to the TWSME. Upon thermal cycling, the depth of the indent can be changed reversibly. Lower indent depth is shown after heating and an increase is shown after cooling.
Based on these results, Zhang et al. proposed a technique to develop a reversible shape memory surface.38-40 They indented NiTi alloys to high strains, thermally cycled
them, and replanarized the samples by stepwise grinding and polishing (Figure 2.7). The multiple grinding and polishing steps were applied in order to remove the remaining indents and regain an optically smooth surface. When heated above Af, the
replanarized surface elicited semispherical surface protrusions at the former positions of the indents. These protrusions were thermally switchable and most likely induced by the residual dislocation structure oriented around the initial indent.
Hence, indentation of the surface leads to a surface form memory by means of a specific indentation and planarization technique. This method (indentation induced TWSME) leads to a thermally reversible surface topography which can be switched between a flat and structured state, depending on the initially applied deformation.
Figure 2.6: The indentation induced two-way shape memory effect. (i.-ii.) Twinned martensite is indented and plastically deformed. (iii.-iv.) The indentation induced stress leads to detwinning and stabilization of martensite variants with a preferred orientation. (v.) If the sample is heated the first time above Af a recovery of the indent takes place. The depth of the initial indent profile
(green) decreases with the austenitic transformation by the associated one-way shape memory effect. (vi.) After cooling the sample below Mf, the indentation depth increases (blue). A reverse
shape change is shown within the martensitic transformation, which is related to the two-way shape memory effect. The profile depth can be changed reversibly by heating (red, v.) and cooling across the transformation temperatures (blue, vi.).
The microstructural mechanisms of the structure formation have not been identified in detail and a characterization of the interacting mechanisms of the structuring phenomena remains unclear. Likewise, the experimental parameters, which may enhance the structure formation and their role during martensite transformation, have not been identified in detail.
Current approaches used either single indentation or scratching to induce the deformation related stress field. The resulting remnant dislocation structure favors the nucleation of certain martensitic variants. The growth of these variants causes a shape change upon cooling and is thus thought to be responsible for the TWSME.39 Based on
this theory, Fei et al. tried to improve the recovery rate of the TWSME by shifting the quasistatic indentation conditions (Hertzian contact at low strain rates) to a dynamical regime by laser shock indentation (high strain rates).41 They observed that the shape
recovery was five times greater in laser shock indentation compared to quasistatic indentation. Accordingly, a higher protrusion size was achieved relative to the deformation depth induced by indentation.
Figure 2.7: Illustration of the preparation steps for switchable surface topographies induced via surface indentation.(A, i) The surface of a NiTi shape memory alloy is plastically deformed by the indenter, leading to a remaining indent. (B) After heating and cooling the sample above and below the phase transformation temperatures a shallower indent profile (blue) remains on the surface due to the two-way shape memory effect. (A, ii) This indent is removed by grinding and subsequent polishing until an optically smooth surface is regained. (A, iii) When the sample is heated above Af, a semispherical protrusion appears in the indented area and disappears again
after cooling below Mf. (C) Thus, by successively heating and cooling the sample, the surface can
2.1.4 Metal forming and structuring processes
The methods of surface texturing for metals range from macroscopic computer numerical controlled machining such as drilling,42 over laser treatments43 and
lithographic technologies,44 to micromachining using focused ion beam technology,45 or
electrochemical processes.46 Preparation methods, in which the topography is formed
serial and point-by-point, are significantly slower compared to parallel texturing, where the whole pattern is transferred at the same time such as in batch or roll-to-roll processes.
Apart from the study by Fei et al., where laser shock indentation was used to prepare switchable TWSME surfaces, all other studies used the same indentation based preparation technique presented by Zhang et al. (Figure 2.7). This technique is a rather complex and time consuming preparation routine.38 Neither single indentation, nor
stepwise planarization is an adequate technique for large-scale production of TWSME surfaces.
In order to select an appropriate machining technology for producing surfaces with switchable topographies on NiTi, the following requirements and specifications have to be considered:
» The size of manufactured structures should range from nano- to millimeters, since usually surface functionalities are related to structures of such a size scale.
» Preferably strong stress fields should be induced highly localized into the surface to promote the TWSME.
» The amount of heat transfer into the work piece should be kept as low as possible in order to avoid any negative influence on the TWSME.
» The process should enable an efficient parallel structuring of larger areas and already be applied or being applicable in industry.
2.1.5 Cold embossing
One of the most direct ways to induce a surface texture into a metal is the surface deformation through embossing or coining. A relatively coarse surface texture with several grooves of 5 to 20 µm in width can be used to improve lubrication for roller/piston in hydraulic motors as shown by Pettersson and Jacobson.47 A specific
rotation to the sliding direction on a steel piston.47-49 With the increasing demand for
miniaturization, the limits of these technologies were currently investigated. Geiger et al. summarized the developments in micro-metalforming and showed that especially the tool manufacturing limits the application for small-scale features.50
Böhm et al. showed that a structure size in the range of a few micrometers can be embossed into a surface with high accuracy (see Figure 2.8). In order to do so, high pressures exceeding the yield strength of the substrate material were applied. The samples were loaded with 2000 N for alumina and more than 5000 N for stainless steel. Subsequently, very high die wear or even failure of the brittle silicon dies was observed.51
Only little research has been carried out regarding the texturing of NiTi shape memory alloys through embossing. Hornbogen described a thermomechanical embossing process as the simplest way to determine the final shape formation in NiTi.52
He clamped the martensitic material in a die and heated it high above Afbut below the
recrystallization temperature. At temperatures above Af the growth of nanometer scale
precipitates, which then affect the martensitic transformation, was favored.
Bradley et al. patented the structuring of a shape memory sheet via embossing using an electromagnetic force-assisted imprint technology.53-54 The features of the
‘embossed shape memory sheet metal article’ should be suitable for the generation of holographic images. The required surface features to gain a visible image is in the range of millimeters and is formed by an electromagnetic forming process. The authors preferred the use of thin foils for this process and applied a thermal treatment in form of heating and cooling during the electromagnetic forming.
Figure 2.8: SEM images of metallic surfaces with complex micro-geometries fabricated via embossing. (A) Silicon die with straight grooves and a gap width of 1 µm. (B) Straight grooves on an alumina (Al99.5) substrate transferred with cold embossing. (C) Silicon die with a complex geometry. D, Complex surface structure transferred by superplastic embossing at 250°C for 12 min.51
Due to the specific stress-strain behavior and high degree of work hardening, in NiTi shape memory alloys, the embossing process leads to a poor structuring accuracy and high tool wear.55 Various studies demonstrated the poor machinability of NiTi
shape memory alloys using conventional machining such as cutting and drilling.55-58
Thus, non-conventional techniques such as electrochemical machining are of growing interest to form NiTi devices.
2.1.6 Pulse electrochemical machining
Pulse electrochemical machining (PECM) is an extension of electrochemical machining (ECM), an electrolysis process without direct contact between tool and sample surface. As shown in Figure 2.9, a negative pattern of a tool is transferred onto a substrate by an anodic dissolution. During the process, a small gap between the sample and the cathode is maintained and flushed with an electrolyte at high flowrates. A high voltage is applied, leading to very high current densities. Heat production is avoided and residues of the manufacturing are removed from the interelectrode gap by a uniform flow of the electrolyte through a flushing chamber.59 The pulse current leads
to higher instant current densities and subsequently enables higher surface quality compared to electrochemical machining with continuous current.
Figure 2.9: Schematic drawing of the PECM process and resulting surface texture. (A) The tool is moved upwards and downwards to widen the interelectrode gap (S, gap distance). A short pulse (I, current pulse) is applied at the lower turning point. Modified after 60. (B) SEM images
of a micro-machined tool and the structured Ni substrate.61 (C) It has yet to be shown if
protruding structures could also be manufactured by using a cathode with cavities of various geometries.
The tool (cathode) is moved in a sinusoidal manner upwards and downwards to widen the interelectrode gap (S, gap distance) between the substrate (anode) for the removal of oxides and trapped air in the electrolyte. A short pulse (I, current pulse) is applied at the lower turning point and induces the anodization of the sample, leading to an increased manufacturing accuracy. Rajurkar et al. mentioned that PECM is already applied to manufacture dies, turbine blades, and precision electronic components with an accuracy of 20 to 100 µm.62-63
Besides the advantages of a residual stress-free structure formation and low production of heat, the geometry of the structures and their accuracy is limited by the electrolytic flow. Hence a sophisticated design of the inverse tool geometry is required.64
In literature, most investigations on nonconventional machining of NiTi alloys focus on electropolishing for a highly accurate surface finish in biomedical applications or electro discharge machining (EDM) to structure implant surfaces of NiTi.65-66 Both
processes do not meet the requirements for the preparation of TWSME topographies, since electropolishing does not allow a formation extending surface structures and EDM, as a thermal process, interacts with the thermo-sensitive TWSME in NiTi. Only few investigations have focused on PECM processing of NiTi.67-68
2.1.7 Applications of two-way shape memory surfaces
In terms of potential applications of TWSME surfaces, little research has been conducted.38 The recovery behavior of the indentation depth upon heating, which can
be described as a self-healing property, has been studied for tribological applications by Ni et al.69 Due to the recovery behavior of the OWSME, former indents and wear scars
faded away upon heating. Because the hardness of NiTi was not high enough, the wear-induced material loss was prevented by coating the surface with a thin film of hard CrN. Using this strategy, wear resistant and self-healing tribological surfaces could be created.
Shaw et al. used the shape memory effect for high-density data storage. Data bits were stored by indentation and deleted by heating, due to the OWSME and the self-healing behavior of indents.70-71
Fei et al. showed that TWSME surfaces could also perform mechanical work with sufficient energy to induce localized plastic deformation in a strong base metal substrate. In planar contact, the formation of protrusions on a NiTi TWSME surface was able to deform stainless steel (304) under compressive loading. Considering that
this process works robustly at the nanoscale, the authors emphasized its possible application in the field of MEMS micro assembly, nanolithography or thermally variable friction surfaces. 40
The previously mentioned studies have focused on binary NiTi alloys, no research has been conducted on further alloys such as ternary or quaternary shape memory alloys. This could lead to different mechanical properties of the surface, enable increased actuation forces or the use of different external stimuli. A transfer of the indentation induced TWSME onto alloys with different phase transformation temperatures would enable an adaptation to the required activation temperatures. For example, an alloy with a phase transformation near body temperature could enable an activation of the TWSME by body heat, leading to new applications in biomedicine. Although not investigated in this thesis, a transfer of the TWSME topographies onto magnetic shape memory alloys could enable switching by a magnetic field.
2.2 Functional microstructures and microtopographies
Apart from the wear resistant coating on a TWSME surface,69 no research has beencarried out combining functional surface structures or other coatings with switchable topographies in NiTi. For many properties, the surface topography at micro- and nanometer scale plays an important role and induces macroscopic effects.72 As
schematically shown in Figure 2.10, functional coatings or micro- and nanostructures could influence effects, such as thermal radiation, wettability, reflectivity or friction and drag reduction.47, 73-77
Figure 2.10: Schematic of different surface properties induced by the interaction of functional surface structures and coatings with external factors. The microtopography plays an important role for macroscopic effects such as (A) heat transfer,73 (B) dry adhesion,78 (C)
superhydrophobicity,74 (D) cell-response,79 (E) anti-reflectivity75-76 and (F) drag80 as well as wear
Microdroplets coupled with specific surface textures were used to promote spreading and evaporation for cooling in high heat flux MEMS.73
Many superhydrophobic surfaces were induced using microstructures mimicking the lotus-effect.81 Lotus plants, possess hierarchically structured waxy crystals on their
leaves, which lead to very little adhesion of water droplets to the surface.74 Also
anti-reflective coatings were mostly inspired by nature and make use of the interaction of light with different microstructures.75-76 For example, nipple arrays, multilayer systems
or microlense arrays in natural photonic structures induce optical effects such as reduced-reflectivity, structural colors or collection of light. 82-84 Other research showed
that friction and drag could be reduced using scales and rib-like microstructures. 85
Such riblet surfaces, which were also found in shark skin, reduce turbulent skin friction if they are aligned parallel to the direction of flow.80 Aircraft drag is shown to be reduced
by up to 8% using a riblet foil.86 Another example from nature using specific
microtopographies to reduce friction is found in the skin of sandfishes and snakes.87-89
Saw-tooth shaped ridges and bumps on the skin lead to a reduced friction coefficient and increased abrasion resistance, along the animals longitudinal axis and higher friction in the opposite direction. Thus, these nano-to microstructures provide friction anisotropy for locomotion. Recently, Greiner and Schäfer showed a reduction of dry sliding forces by more than 40% using such a bioinspired morphological surface texture.89 They anticipate that these results could have a significant impact on all
applications with dry sliding contacts. Other microstructures and topographies with a specific morphology were used to induce adhesive properties or to control cell-alignment on biomedical surfaces.79, 90-91
Accordingly, the combination of such functional surface structures with a TWSME surface or substrate material would enable control and switch the specific properties related to the surface structures by temperature.
2.2.1 Bioinspired dry adhesives and switchable adhesion
The ability of a surface to adhere to a substrate depends on chemical interaction, proximity and area of contact. The latter two are strongly influenced by surface topography. Inspired by the attachment devices of natural prototypes, such as spiders and geckos (see Figure 2.11), dry adhesives usually use functional microstructures to enhance van der Waals interactions in order to get robust adhesion on various surfaces.92
By the use of fibrillar structures, the contact area is split into many single contacts, the short-range Van-der-Waals forces are enhanced and adhesion is increased.93
Kamperman et al. summarized different mechanisms interacting with the topography of the substrate to promote dry-adhesion by contact splitting (Figure 2.12).90
The hierarchical fibrils on the micro- and nanometer scale allow conformal contact with the substrate to increase adhesion.93-94 The flexibility of the fibrillar structures
enables enhanced adaptability to smooth, but also to rough, substrates.95 Furthermore,
the high number of separated contact elements leads to higher defect tolerance and redundancy.96
Figure 2.11: Terminal elements (circles) of hairy attachment pads in animals of different size. With increasing body weight, finer fibrillary structures to enhance adhesion are shown.93
Figure 2.12: Schematic illustration of mechanisms, which enhance adhesion by contact splitting in fibrillary dry adhesive surfaces. (A) The reduced radius and increased number of contact elements induced a size effect enhancing the short-range van der Waals forces. (B) highly flexible features enable adaptability to rough topographies. (C) Increase in work of adhesion due to discontinuous crack propagation. (D) low susceptibility to interfacial defects. Modified after Kamperman et al. 90
In the last decade, many different surface structures were developed as synthetic dry adhesives to benefit from nature’s contact-splitting principle.90, 97-98 Greiner et al.
performed a systematic study on elastomeric pillars with a variation in aspect-ratio, tip shape, and backing layer thickness.99-100 Fibrillar structures terminated with flanged
ends, so called mushroom tips, received particular interest due to their strong adhesive behavior. The detachment of such bioinspired microstructures usually requires peeling or shear movements, which are difficult to adapt to industrial fabrication processes.
For industrial applications, the release from the substrate by detachment of the dry adhesive structures is as important as the increase in adhesion, to allow attachment. The ability to attach to, and detach from various surfaces without leaving residues or causing any mechanical damages is highly beneficial for many industry processes.
For example, in pick-and-place processes and for the handling of sensitive products, a reliable short time attachment is required. Thus, more recently, switchable adhesive systems have been developed.101-106
In many of these studies a change in topography leads to an increase or decrease in contact area, and subsequently, to a change in the adhesive performance (Figure 2.13). Northen et al. combined an adhesive polymer with magneto-sensitive Ni-cantilevers for reversible adhesion (Figure 2.13, A).101 They applied a magnetic field to change the
orientation of the cantilevers that induced a change in contact area and subsequently in adhesion. Their system showed excellent switchability, but the complicated fabrication process and the low adhesive strength limited its application. Reddy et al. used a temperature triggered thermoplastic elastomer in order to change the tilting angle of fibrils (Figure 2.13, B). Mechanical tilting of these structures led to an adhesion loss, and subsequent heating recovered the original orientation of the structures and by this the adhesive state.102 A one-time switch in adhesion was possible, but no
reversibility of the effect could be obtained. Kim et al. developed a mechanism based on a temporary, metastable contact for picking up and releasing small objects (Figure 2.13, C).107 This system lacked active control in adhesion, which significantly limits the
process reliability and maximum weight of the objects. Paretkar et al. presented a reversible pressure actuated adhesive system (Figure 2.13, D).103 They triggered the
loss in adhesion by mechanical loading, causing a reversible buckling and subsequent loss of intimate contact of the structures with the testing surface. Accordingly, when a threshold compressive load was applied to the fibrillar system, adhesion dropped significantly. In comparison to cylindrical structures, the mushroom structures
required higher preload to force buckling. This is plausible, because their flanged ends increase contact adaptability.108
Other systems were triggered by light or humidity.109-110 Boyne et al. found a
reduction of peel strength dependent on light intensity and irradiation time in a pressure sensitive adhesive.109 Humidity triggered the adhesion in porous polymer pads
presented by Xue et al.110 In the sponge-like structures adhesion increased due to a
humidity-induced decrease in the elastic modulus of the polymer.
Figure 2.13: Overview on switchable adhesive systems. (A) Freestanding nickel cantilevers with paddles of nanorods for switchable adhesion stimulated using a magnetic field.101 (B) Shape
memory polymer pillars were bend and triggered by temperature to change contact area. (C) Elastomeric micropyramidal structures for use in assembly by transfer printing.111 (D)
Most approaches in switchable adhesion induce a change of the adhesive structures themselves. Only few took advantage to change the shape or topography of the backing layer in order to switch adhesion (Figure 2.14).105-106
Xie et al. induced a peeling motion, analogue to the detachment movement in gecko toes, in a double layer system of a shape memory polymer with an elastomeric adhesive polymer. Due to a slight curvature of the shape memory polymer, the compliance to the substrate was reduced and the sample detached.
Jeong et al. presented a similar approach. They used the formation of wrinkles on a polymer sheet to misalign fibrillar adhesive structures positioned on top of the sheet.106
By stretching the backing layer, the fibrils oriented in a normal direction to the surface of contact and, by relaxing the backing layer, the wrinkled topography led to misalignment with respect to the surface of contact. Thus, a control of adhesion in normal and shear direction was achieved.
2.2.2 Influence of (active) topographies on cells
The interface between tissue and an implant surface plays a crucial role during the integration and healing process in a body. Therefore, the cell-surface interaction of implant materials and different cell cultures has been investigated for years. Curtis and Varde showed, already in 1964, how the topography of an implant surface controls the interaction with cells and contributes positively to tissue repair.113
Figure 2.14: Two examples for switchable adhesives using a topographic change of the backing layer to control the area of contact to the substrate surface. (A) A self-peeling dry adhesive is build using a two layer composite with a dry adhesive layer and a shape memory polymer.105 (B)
Surface wrinkling was controlled by an applied strain. In the stretched state the pillars are in contact and adhesion increases, in the released state the pillars tilt and adhesion is reduced.106
